adapters-opt-bnb8-QLORA-super_glue-wsc
/
trainer_state-opt-bnb8-QLORA-super_glue-wsc-sequence_classification.json
| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 1, | |
| "global_step": 70, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 32.66877746582031, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.2134, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "eval_accuracy": 0.38636363636363635, | |
| "eval_loss": 1.4526692628860474, | |
| "eval_runtime": 1.8581, | |
| "eval_samples_per_second": 71.041, | |
| "eval_steps_per_second": 2.153, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 36.16679382324219, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3108, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "eval_accuracy": 0.4015151515151515, | |
| "eval_loss": 1.356563687324524, | |
| "eval_runtime": 1.8456, | |
| "eval_samples_per_second": 71.522, | |
| "eval_steps_per_second": 2.167, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 33.828529357910156, | |
| "learning_rate": 4.9264705882352944e-05, | |
| "loss": 1.2273, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "eval_accuracy": 0.42424242424242425, | |
| "eval_loss": 1.1141084432601929, | |
| "eval_runtime": 1.8499, | |
| "eval_samples_per_second": 71.355, | |
| "eval_steps_per_second": 2.162, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 30.754159927368164, | |
| "learning_rate": 4.8529411764705885e-05, | |
| "loss": 1.0903, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "eval_accuracy": 0.4166666666666667, | |
| "eval_loss": 0.9176180362701416, | |
| "eval_runtime": 1.8506, | |
| "eval_samples_per_second": 71.328, | |
| "eval_steps_per_second": 2.161, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 17.295827865600586, | |
| "learning_rate": 4.7794117647058826e-05, | |
| "loss": 0.8305, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "eval_accuracy": 0.4166666666666667, | |
| "eval_loss": 0.7779105305671692, | |
| "eval_runtime": 1.8478, | |
| "eval_samples_per_second": 71.435, | |
| "eval_steps_per_second": 2.165, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 10.875761985778809, | |
| "learning_rate": 4.705882352941177e-05, | |
| "loss": 0.7782, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "eval_accuracy": 0.6212121212121212, | |
| "eval_loss": 0.6866840124130249, | |
| "eval_runtime": 1.8525, | |
| "eval_samples_per_second": 71.256, | |
| "eval_steps_per_second": 2.159, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.133961200714111, | |
| "learning_rate": 4.632352941176471e-05, | |
| "loss": 0.7671, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6212121212121212, | |
| "eval_loss": 0.6688076257705688, | |
| "eval_runtime": 1.85, | |
| "eval_samples_per_second": 71.353, | |
| "eval_steps_per_second": 2.162, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 16.225610733032227, | |
| "learning_rate": 4.558823529411765e-05, | |
| "loss": 0.8476, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "eval_accuracy": 0.6287878787878788, | |
| "eval_loss": 0.663224995136261, | |
| "eval_runtime": 1.8537, | |
| "eval_samples_per_second": 71.21, | |
| "eval_steps_per_second": 2.158, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 8.729240417480469, | |
| "learning_rate": 4.485294117647059e-05, | |
| "loss": 0.7226, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "eval_accuracy": 0.5984848484848485, | |
| "eval_loss": 0.6746981739997864, | |
| "eval_runtime": 1.8561, | |
| "eval_samples_per_second": 71.118, | |
| "eval_steps_per_second": 2.155, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 4.27821159362793, | |
| "learning_rate": 4.411764705882353e-05, | |
| "loss": 0.6843, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "eval_accuracy": 0.5984848484848485, | |
| "eval_loss": 0.6854373812675476, | |
| "eval_runtime": 1.8512, | |
| "eval_samples_per_second": 71.304, | |
| "eval_steps_per_second": 2.161, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 5.028013229370117, | |
| "learning_rate": 4.3382352941176474e-05, | |
| "loss": 0.6514, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "eval_accuracy": 0.5681818181818182, | |
| "eval_loss": 0.7006022334098816, | |
| "eval_runtime": 1.8521, | |
| "eval_samples_per_second": 71.269, | |
| "eval_steps_per_second": 2.16, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 4.5469865798950195, | |
| "learning_rate": 4.2647058823529415e-05, | |
| "loss": 0.7274, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.7056108117103577, | |
| "eval_runtime": 1.8545, | |
| "eval_samples_per_second": 71.18, | |
| "eval_steps_per_second": 2.157, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 3.2658820152282715, | |
| "learning_rate": 4.1911764705882356e-05, | |
| "loss": 0.7121, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.7070386409759521, | |
| "eval_runtime": 1.8579, | |
| "eval_samples_per_second": 71.048, | |
| "eval_steps_per_second": 2.153, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 4.191383361816406, | |
| "learning_rate": 4.11764705882353e-05, | |
| "loss": 0.7359, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.7112393379211426, | |
| "eval_runtime": 1.8518, | |
| "eval_samples_per_second": 71.28, | |
| "eval_steps_per_second": 2.16, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 3.1839404106140137, | |
| "learning_rate": 4.044117647058824e-05, | |
| "loss": 0.6841, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.7038382291793823, | |
| "eval_runtime": 1.8491, | |
| "eval_samples_per_second": 71.386, | |
| "eval_steps_per_second": 2.163, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 3.6612770557403564, | |
| "learning_rate": 3.970588235294117e-05, | |
| "loss": 0.7365, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.6875917911529541, | |
| "eval_runtime": 1.8532, | |
| "eval_samples_per_second": 71.229, | |
| "eval_steps_per_second": 2.158, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "grad_norm": 11.7686128616333, | |
| "learning_rate": 3.897058823529412e-05, | |
| "loss": 0.7549, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "eval_accuracy": 0.5909090909090909, | |
| "eval_loss": 0.6774886846542358, | |
| "eval_runtime": 1.8522, | |
| "eval_samples_per_second": 71.266, | |
| "eval_steps_per_second": 2.16, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 6.852344512939453, | |
| "learning_rate": 3.8235294117647055e-05, | |
| "loss": 0.7206, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "eval_accuracy": 0.5984848484848485, | |
| "eval_loss": 0.6746863126754761, | |
| "eval_runtime": 1.8521, | |
| "eval_samples_per_second": 71.27, | |
| "eval_steps_per_second": 2.16, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 2.1575918197631836, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.7058, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "eval_accuracy": 0.6060606060606061, | |
| "eval_loss": 0.6730439066886902, | |
| "eval_runtime": 1.8571, | |
| "eval_samples_per_second": 71.078, | |
| "eval_steps_per_second": 2.154, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 4.095081329345703, | |
| "learning_rate": 3.6764705882352945e-05, | |
| "loss": 0.7035, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "eval_accuracy": 0.6060606060606061, | |
| "eval_loss": 0.6729995012283325, | |
| "eval_runtime": 1.8482, | |
| "eval_samples_per_second": 71.419, | |
| "eval_steps_per_second": 2.164, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4.262673854827881, | |
| "learning_rate": 3.6029411764705886e-05, | |
| "loss": 0.7604, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6060606060606061, | |
| "eval_loss": 0.6814305782318115, | |
| "eval_runtime": 1.8497, | |
| "eval_samples_per_second": 71.361, | |
| "eval_steps_per_second": 2.162, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 3.142857142857143, | |
| "grad_norm": 2.6779417991638184, | |
| "learning_rate": 3.529411764705883e-05, | |
| "loss": 0.7304, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 3.142857142857143, | |
| "eval_accuracy": 0.553030303030303, | |
| "eval_loss": 0.6932557821273804, | |
| "eval_runtime": 1.8503, | |
| "eval_samples_per_second": 71.34, | |
| "eval_steps_per_second": 2.162, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 3.2857142857142856, | |
| "grad_norm": 2.625526189804077, | |
| "learning_rate": 3.455882352941177e-05, | |
| "loss": 0.7007, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 3.2857142857142856, | |
| "eval_accuracy": 0.42424242424242425, | |
| "eval_loss": 0.707717776298523, | |
| "eval_runtime": 1.8519, | |
| "eval_samples_per_second": 71.279, | |
| "eval_steps_per_second": 2.16, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 6.1237053871154785, | |
| "learning_rate": 3.382352941176471e-05, | |
| "loss": 0.7072, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "eval_accuracy": 0.38636363636363635, | |
| "eval_loss": 0.7181107997894287, | |
| "eval_runtime": 1.8622, | |
| "eval_samples_per_second": 70.883, | |
| "eval_steps_per_second": 2.148, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 6.0173659324646, | |
| "learning_rate": 3.308823529411765e-05, | |
| "loss": 0.7625, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "eval_accuracy": 0.3787878787878788, | |
| "eval_loss": 0.7258760333061218, | |
| "eval_runtime": 1.847, | |
| "eval_samples_per_second": 71.469, | |
| "eval_steps_per_second": 2.166, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.7142857142857144, | |
| "grad_norm": 3.7744202613830566, | |
| "learning_rate": 3.235294117647059e-05, | |
| "loss": 0.6608, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.7142857142857144, | |
| "eval_accuracy": 0.38636363636363635, | |
| "eval_loss": 0.7310694456100464, | |
| "eval_runtime": 1.8004, | |
| "eval_samples_per_second": 73.317, | |
| "eval_steps_per_second": 2.222, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.857142857142857, | |
| "grad_norm": 6.532558917999268, | |
| "learning_rate": 3.161764705882353e-05, | |
| "loss": 0.6962, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 3.857142857142857, | |
| "eval_accuracy": 0.3787878787878788, | |
| "eval_loss": 0.7323981523513794, | |
| "eval_runtime": 1.8509, | |
| "eval_samples_per_second": 71.317, | |
| "eval_steps_per_second": 2.161, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 7.10076379776001, | |
| "learning_rate": 3.0882352941176475e-05, | |
| "loss": 0.7431, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.3712121212121212, | |
| "eval_loss": 0.7279858589172363, | |
| "eval_runtime": 1.8496, | |
| "eval_samples_per_second": 71.367, | |
| "eval_steps_per_second": 2.163, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.142857142857143, | |
| "grad_norm": 6.7147088050842285, | |
| "learning_rate": 3.0147058823529413e-05, | |
| "loss": 0.7184, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 4.142857142857143, | |
| "eval_accuracy": 0.3787878787878788, | |
| "eval_loss": 0.7212890982627869, | |
| "eval_runtime": 1.853, | |
| "eval_samples_per_second": 71.238, | |
| "eval_steps_per_second": 2.159, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 8.177295684814453, | |
| "learning_rate": 2.9411764705882354e-05, | |
| "loss": 0.6988, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "eval_accuracy": 0.5075757575757576, | |
| "eval_loss": 0.7096087336540222, | |
| "eval_runtime": 1.8491, | |
| "eval_samples_per_second": 71.385, | |
| "eval_steps_per_second": 2.163, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.428571428571429, | |
| "grad_norm": 3.7088816165924072, | |
| "learning_rate": 2.8676470588235295e-05, | |
| "loss": 0.6843, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 4.428571428571429, | |
| "eval_accuracy": 0.553030303030303, | |
| "eval_loss": 0.7047467231750488, | |
| "eval_runtime": 1.8517, | |
| "eval_samples_per_second": 71.286, | |
| "eval_steps_per_second": 2.16, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 2.9432482719421387, | |
| "learning_rate": 2.7941176470588236e-05, | |
| "loss": 0.7146, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.6983250975608826, | |
| "eval_runtime": 1.8506, | |
| "eval_samples_per_second": 71.328, | |
| "eval_steps_per_second": 2.161, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.714285714285714, | |
| "grad_norm": 3.511507511138916, | |
| "learning_rate": 2.7205882352941174e-05, | |
| "loss": 0.7224, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 4.714285714285714, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.6929479837417603, | |
| "eval_runtime": 1.8475, | |
| "eval_samples_per_second": 71.448, | |
| "eval_steps_per_second": 2.165, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "grad_norm": 4.023740291595459, | |
| "learning_rate": 2.647058823529412e-05, | |
| "loss": 0.731, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.6917376518249512, | |
| "eval_runtime": 1.85, | |
| "eval_samples_per_second": 71.351, | |
| "eval_steps_per_second": 2.162, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.271909475326538, | |
| "learning_rate": 2.5735294117647057e-05, | |
| "loss": 0.6889, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.6847301125526428, | |
| "eval_runtime": 1.8509, | |
| "eval_samples_per_second": 71.318, | |
| "eval_steps_per_second": 2.161, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 5.142857142857143, | |
| "grad_norm": 4.620840072631836, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.751, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 5.142857142857143, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.6826764345169067, | |
| "eval_runtime": 1.8504, | |
| "eval_samples_per_second": 71.335, | |
| "eval_steps_per_second": 2.162, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 5.285714285714286, | |
| "grad_norm": 8.111146926879883, | |
| "learning_rate": 2.4264705882352942e-05, | |
| "loss": 0.7243, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 5.285714285714286, | |
| "eval_accuracy": 0.5833333333333334, | |
| "eval_loss": 0.6787553429603577, | |
| "eval_runtime": 1.8504, | |
| "eval_samples_per_second": 71.336, | |
| "eval_steps_per_second": 2.162, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 5.428571428571429, | |
| "grad_norm": 6.502111911773682, | |
| "learning_rate": 2.3529411764705884e-05, | |
| "loss": 0.7345, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 5.428571428571429, | |
| "eval_accuracy": 0.5984848484848485, | |
| "eval_loss": 0.6770064830780029, | |
| "eval_runtime": 1.8504, | |
| "eval_samples_per_second": 71.336, | |
| "eval_steps_per_second": 2.162, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 5.571428571428571, | |
| "grad_norm": 2.404085397720337, | |
| "learning_rate": 2.2794117647058825e-05, | |
| "loss": 0.7172, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 5.571428571428571, | |
| "eval_accuracy": 0.6060606060606061, | |
| "eval_loss": 0.6782048344612122, | |
| "eval_runtime": 1.8536, | |
| "eval_samples_per_second": 71.211, | |
| "eval_steps_per_second": 2.158, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 3.698408842086792, | |
| "learning_rate": 2.2058823529411766e-05, | |
| "loss": 0.6653, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "eval_accuracy": 0.5984848484848485, | |
| "eval_loss": 0.6809452176094055, | |
| "eval_runtime": 1.853, | |
| "eval_samples_per_second": 71.234, | |
| "eval_steps_per_second": 2.159, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.857142857142857, | |
| "grad_norm": 4.651946067810059, | |
| "learning_rate": 2.1323529411764707e-05, | |
| "loss": 0.7222, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 5.857142857142857, | |
| "eval_accuracy": 0.5984848484848485, | |
| "eval_loss": 0.6832801103591919, | |
| "eval_runtime": 1.8003, | |
| "eval_samples_per_second": 73.321, | |
| "eval_steps_per_second": 2.222, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.0304791927337646, | |
| "learning_rate": 2.058823529411765e-05, | |
| "loss": 0.6938, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5909090909090909, | |
| "eval_loss": 0.6841915845870972, | |
| "eval_runtime": 1.8506, | |
| "eval_samples_per_second": 71.329, | |
| "eval_steps_per_second": 2.161, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 6.142857142857143, | |
| "grad_norm": 2.409463882446289, | |
| "learning_rate": 1.9852941176470586e-05, | |
| "loss": 0.721, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 6.142857142857143, | |
| "eval_accuracy": 0.5984848484848485, | |
| "eval_loss": 0.6831972599029541, | |
| "eval_runtime": 1.8507, | |
| "eval_samples_per_second": 71.326, | |
| "eval_steps_per_second": 2.161, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 6.285714285714286, | |
| "grad_norm": 6.6015305519104, | |
| "learning_rate": 1.9117647058823528e-05, | |
| "loss": 0.6961, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 6.285714285714286, | |
| "eval_accuracy": 0.5909090909090909, | |
| "eval_loss": 0.6853367686271667, | |
| "eval_runtime": 1.8482, | |
| "eval_samples_per_second": 71.421, | |
| "eval_steps_per_second": 2.164, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "grad_norm": 2.793956756591797, | |
| "learning_rate": 1.8382352941176472e-05, | |
| "loss": 0.7025, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "eval_accuracy": 0.5681818181818182, | |
| "eval_loss": 0.687698245048523, | |
| "eval_runtime": 1.8519, | |
| "eval_samples_per_second": 71.277, | |
| "eval_steps_per_second": 2.16, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 6.571428571428571, | |
| "grad_norm": 4.182642936706543, | |
| "learning_rate": 1.7647058823529414e-05, | |
| "loss": 0.7034, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 6.571428571428571, | |
| "eval_accuracy": 0.553030303030303, | |
| "eval_loss": 0.6860617995262146, | |
| "eval_runtime": 1.8526, | |
| "eval_samples_per_second": 71.251, | |
| "eval_steps_per_second": 2.159, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 6.714285714285714, | |
| "grad_norm": 6.431582450866699, | |
| "learning_rate": 1.6911764705882355e-05, | |
| "loss": 0.7481, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 6.714285714285714, | |
| "eval_accuracy": 0.5606060606060606, | |
| "eval_loss": 0.688340425491333, | |
| "eval_runtime": 1.8572, | |
| "eval_samples_per_second": 71.073, | |
| "eval_steps_per_second": 2.154, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "grad_norm": 6.1300811767578125, | |
| "learning_rate": 1.6176470588235296e-05, | |
| "loss": 0.7204, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "eval_accuracy": 0.5227272727272727, | |
| "eval_loss": 0.6901603937149048, | |
| "eval_runtime": 1.8531, | |
| "eval_samples_per_second": 71.231, | |
| "eval_steps_per_second": 2.159, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.4620742797851562, | |
| "learning_rate": 1.5441176470588237e-05, | |
| "loss": 0.7339, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5757575757575758, | |
| "eval_loss": 0.6880089640617371, | |
| "eval_runtime": 1.8595, | |
| "eval_samples_per_second": 70.985, | |
| "eval_steps_per_second": 2.151, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 4.01107120513916, | |
| "learning_rate": 1.4705882352941177e-05, | |
| "loss": 0.739, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "eval_accuracy": 0.553030303030303, | |
| "eval_loss": 0.6889382004737854, | |
| "eval_runtime": 1.8607, | |
| "eval_samples_per_second": 70.94, | |
| "eval_steps_per_second": 2.15, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.285714285714286, | |
| "grad_norm": 1.751711368560791, | |
| "learning_rate": 1.3970588235294118e-05, | |
| "loss": 0.7146, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 7.285714285714286, | |
| "eval_accuracy": 0.5, | |
| "eval_loss": 0.6900361180305481, | |
| "eval_runtime": 1.8585, | |
| "eval_samples_per_second": 71.024, | |
| "eval_steps_per_second": 2.152, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 7.428571428571429, | |
| "grad_norm": 2.6621031761169434, | |
| "learning_rate": 1.323529411764706e-05, | |
| "loss": 0.7117, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 7.428571428571429, | |
| "eval_accuracy": 0.48484848484848486, | |
| "eval_loss": 0.6922881603240967, | |
| "eval_runtime": 1.8588, | |
| "eval_samples_per_second": 71.013, | |
| "eval_steps_per_second": 2.152, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 7.571428571428571, | |
| "grad_norm": 9.999210357666016, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.739, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 7.571428571428571, | |
| "eval_accuracy": 0.4772727272727273, | |
| "eval_loss": 0.6930486559867859, | |
| "eval_runtime": 1.8501, | |
| "eval_samples_per_second": 71.347, | |
| "eval_steps_per_second": 2.162, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 7.714285714285714, | |
| "grad_norm": 3.8620285987854004, | |
| "learning_rate": 1.1764705882352942e-05, | |
| "loss": 0.6656, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 7.714285714285714, | |
| "eval_accuracy": 0.48484848484848486, | |
| "eval_loss": 0.6901634335517883, | |
| "eval_runtime": 1.8622, | |
| "eval_samples_per_second": 70.885, | |
| "eval_steps_per_second": 2.148, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 6.469612121582031, | |
| "learning_rate": 1.1029411764705883e-05, | |
| "loss": 0.734, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "eval_accuracy": 0.5378787878787878, | |
| "eval_loss": 0.689450204372406, | |
| "eval_runtime": 1.8503, | |
| "eval_samples_per_second": 71.338, | |
| "eval_steps_per_second": 2.162, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.389843225479126, | |
| "learning_rate": 1.0294117647058824e-05, | |
| "loss": 0.7498, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5303030303030303, | |
| "eval_loss": 0.6898378133773804, | |
| "eval_runtime": 1.8571, | |
| "eval_samples_per_second": 71.078, | |
| "eval_steps_per_second": 2.154, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 8.142857142857142, | |
| "grad_norm": 2.5719523429870605, | |
| "learning_rate": 9.558823529411764e-06, | |
| "loss": 0.7198, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 8.142857142857142, | |
| "eval_accuracy": 0.5454545454545454, | |
| "eval_loss": 0.6896809935569763, | |
| "eval_runtime": 1.852, | |
| "eval_samples_per_second": 71.276, | |
| "eval_steps_per_second": 2.16, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 8.285714285714286, | |
| "grad_norm": 3.1081857681274414, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 0.6897, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 8.285714285714286, | |
| "eval_accuracy": 0.5454545454545454, | |
| "eval_loss": 0.6905273795127869, | |
| "eval_runtime": 1.8553, | |
| "eval_samples_per_second": 71.146, | |
| "eval_steps_per_second": 2.156, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 8.428571428571429, | |
| "grad_norm": 2.6777215003967285, | |
| "learning_rate": 8.088235294117648e-06, | |
| "loss": 0.7043, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 8.428571428571429, | |
| "eval_accuracy": 0.5, | |
| "eval_loss": 0.691486120223999, | |
| "eval_runtime": 1.8524, | |
| "eval_samples_per_second": 71.259, | |
| "eval_steps_per_second": 2.159, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 4.99827766418457, | |
| "learning_rate": 7.3529411764705884e-06, | |
| "loss": 0.7178, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "eval_accuracy": 0.5303030303030303, | |
| "eval_loss": 0.6908735632896423, | |
| "eval_runtime": 1.8491, | |
| "eval_samples_per_second": 71.384, | |
| "eval_steps_per_second": 2.163, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 8.714285714285714, | |
| "grad_norm": 9.642988204956055, | |
| "learning_rate": 6.61764705882353e-06, | |
| "loss": 0.682, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 8.714285714285714, | |
| "eval_accuracy": 0.5151515151515151, | |
| "eval_loss": 0.690370500087738, | |
| "eval_runtime": 1.8482, | |
| "eval_samples_per_second": 71.421, | |
| "eval_steps_per_second": 2.164, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 8.857142857142858, | |
| "grad_norm": 6.172641754150391, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.7256, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 8.857142857142858, | |
| "eval_accuracy": 0.5454545454545454, | |
| "eval_loss": 0.6899236440658569, | |
| "eval_runtime": 1.8487, | |
| "eval_samples_per_second": 71.401, | |
| "eval_steps_per_second": 2.164, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 8.340130805969238, | |
| "learning_rate": 5.147058823529412e-06, | |
| "loss": 0.7138, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5303030303030303, | |
| "eval_loss": 0.6913559436798096, | |
| "eval_runtime": 1.8495, | |
| "eval_samples_per_second": 71.372, | |
| "eval_steps_per_second": 2.163, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "grad_norm": 5.535020351409912, | |
| "learning_rate": 4.411764705882353e-06, | |
| "loss": 0.7386, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "eval_accuracy": 0.5, | |
| "eval_loss": 0.6916341185569763, | |
| "eval_runtime": 1.8504, | |
| "eval_samples_per_second": 71.337, | |
| "eval_steps_per_second": 2.162, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "grad_norm": 5.427670001983643, | |
| "learning_rate": 3.6764705882352942e-06, | |
| "loss": 0.7017, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "eval_accuracy": 0.5454545454545454, | |
| "eval_loss": 0.691832423210144, | |
| "eval_runtime": 1.8498, | |
| "eval_samples_per_second": 71.358, | |
| "eval_steps_per_second": 2.162, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 9.428571428571429, | |
| "grad_norm": 3.197458505630493, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 0.7449, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 9.428571428571429, | |
| "eval_accuracy": 0.5378787878787878, | |
| "eval_loss": 0.6921135187149048, | |
| "eval_runtime": 1.8493, | |
| "eval_samples_per_second": 71.38, | |
| "eval_steps_per_second": 2.163, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 9.571428571428571, | |
| "grad_norm": 1.1438813209533691, | |
| "learning_rate": 2.2058823529411767e-06, | |
| "loss": 0.6833, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 9.571428571428571, | |
| "eval_accuracy": 0.5454545454545454, | |
| "eval_loss": 0.6940193176269531, | |
| "eval_runtime": 1.8523, | |
| "eval_samples_per_second": 71.265, | |
| "eval_steps_per_second": 2.16, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 9.714285714285714, | |
| "grad_norm": 6.748230934143066, | |
| "learning_rate": 1.4705882352941177e-06, | |
| "loss": 0.7226, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 9.714285714285714, | |
| "eval_accuracy": 0.553030303030303, | |
| "eval_loss": 0.6915127038955688, | |
| "eval_runtime": 1.8475, | |
| "eval_samples_per_second": 71.45, | |
| "eval_steps_per_second": 2.165, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 9.857142857142858, | |
| "grad_norm": 2.6640241146087646, | |
| "learning_rate": 7.352941176470589e-07, | |
| "loss": 0.6703, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 9.857142857142858, | |
| "eval_accuracy": 0.5227272727272727, | |
| "eval_loss": 0.691817581653595, | |
| "eval_runtime": 1.8496, | |
| "eval_samples_per_second": 71.367, | |
| "eval_steps_per_second": 2.163, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 5.097612380981445, | |
| "learning_rate": 0.0, | |
| "loss": 0.6853, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5075757575757576, | |
| "eval_loss": 0.6922229528427124, | |
| "eval_runtime": 1.8465, | |
| "eval_samples_per_second": 71.487, | |
| "eval_steps_per_second": 2.166, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 70, | |
| "total_flos": 2.095386656033997e+16, | |
| "train_loss": 0.7473028591700963, | |
| "train_runtime": 438.8611, | |
| "train_samples_per_second": 11.986, | |
| "train_steps_per_second": 0.16 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 70, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2.095386656033997e+16, | |
| "train_batch_size": 10, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |